Treatment outcome in patients with neuropathic pain (NP) is often variable and disappointing. We tested the hypothesis that patients with clear evidence of nervous system lesion respond better to pharmacological treatment with documented effect on NP than patients with poor or no evidence of nervous system lesion. Furthermore, we examined whether specific symptoms or signs were associated with treatment outcome. A total of 214 patients with suspected non-cancer NP were divided into four groups with graded evidence of nervous system lesion based on medical history, bedside sensory examination, quantitative sensory tests, electrophysiology, and neuroimaging. Patients were treated with imipramine guided by plasma-drug concentrations. Gabapentin 2400 mg/day was given in case of treatment failure or if imipramine treatment was not possible. Two hundred patients completed the study. Global pain relief was similar in the four groups. There was no association between evidence of nervous system lesion and treatment outcome. Classical NP signs: abnormal temporal summation, cold and brush allodynia, and abnormal sensibility to temperature were also unrelated to outcome. Treatment outcome was similar in peripheral and central definite NP. Neither definite evidence of nervous system lesion nor abnormal sensory phenomena seems to predict for good outcome of therapy with imipramine or gabapentin in patients with suspected neuropathic pain.